Hearing record

Day 20

Evidence led by Adv Matthew Chaskalson SC

Witness: Brigadier Mishack Mkhabela. The SAPS ballistics head faced sustained questioning over errors and omissions in a forensic report tied to the murder of engineer Armand Swart, rejecting suggestions of sabotage and blaming 'human error' while laying bare a forensics unit buckling under tens of thousands of backlogged cases.

Still from Day 20 of the Madlanga Commission

The day in brief

What you need to know

  • Day 20 (27 October 2025) was led by Brigadier Mishack Mkhabela, head of SAPS ballistics, testifying on a disputed forensic report from the Pretoria Forensic Science Laboratory in Silverton.
  • The report, compiled by Captain Itumeleng Makgotloe, was flagged by detectives for gaps: no mention of 15 AK-47 cartridges, mismatched case numbers, wrong dates, and a failure to link firearms to other crimes.
  • Chief evidence leader Adv Matthew Chaskalson SC put it to Mkhabela that a wrong case number amounted to a deliberate attempt to sabotage the prosecution; Mkhabela rejected this, calling the problems typing and human error.
  • Mkhabela presented evidence that four firearms from the Bramley haul, recovered after Swart's murder, were linked to 24 other violent crimes, including at least 20 murders.
  • He testified that SAPS had repurposed a museum into secure storage for more than 29,000 firearms linked to killings nationwide.
  • He disclosed a ballistics backlog of more than 41,000 criminal cases, alongside severe staff shortages crippling firearm investigations.
  • Mkhabela is not accused of a crime; his evidence describes systemic failures and rejects allegations of deliberate tampering.

Full coverage

The day, in depth

'Human error', not sabotage

Mkhabela pushed back hard against the suggestion that flaws in the Silverton ballistics report were deliberate. Where detectives saw tampering that could hand defence lawyers an argument that no link existed to the accused, he described omissions that were caught and fixed. TimesLIVE and Central News reported his insistence that the errors were administrative rather than an attempt to derail the case against those arrested over Armand Swart's killing.

A forensics system under strain

Beyond the single report, Mkhabela's evidence sketched a ballistics capability close to breaking point. He told the commission of a backlog exceeding 41,000 cases, chronic understaffing, and a stockpile of more than 29,000 firearms so large that a museum had been converted into secure storage. IOL and EWN reported his account of delays rippling across firearm investigations nationwide.

One haul, many killings

Mkhabela linked four firearms recovered in Bramley after the Swart murder to 24 other violent crimes, including at least 20 murders, illustrating how a single cache of weapons can thread through a web of unrelated cases and why accurate ballistics linkage matters so much to prosecutions.

In their words

Key moments from the record

The omission was identified, investigated, and corrected. There was no manipulation of evidence.
Sabotage rejected. Responding to Adv Matthew Chaskalson SC's suggestion that a wrong case number in the ballistics report was a deliberate attempt to sabotage the prosecution.
The same firearms recovered in Bramley following Swart's murder were connected to over 20 other murders.
One haul, twenty murders. Presenting a graphic linking firearms recovered in Bramley after Armand Swart's murder to a string of other cases.

Our analysis

The bottom line

The commission's ballistics turn shifts the inquiry's centre of gravity from intent to competence. Where earlier testimony probed whether officials protected wrongdoers, Mkhabela's evidence asks a quieter but no less damaging question: whether the state's forensic machinery is capable of proving anything at all. A 41,000-case backlog and a museum full of unlogged firearms are not scandals of corruption so much as of capacity, and the two can be hard to tell apart from the outside.

That ambiguity is exactly what the sabotage-versus-error dispute exposes. Chaskalson's framing matters because a single wrong case number can collapse a prosecution whether it was planted or fat-fingered, and the accused benefit either way. Mkhabela's defence, that the omission was found and corrected, is plausible on its face, but it also concedes the report left the lab flawed. Fairly read, his testimony is less an exoneration of the system than a portrait of one so overloaded that honest mistakes and deliberate ones become indistinguishable, which is its own kind of vulnerability for the justice system to carry.

None of this pronounces on guilt. Mkhabela is a witness describing his unit's work, not an accused, and the detectives' tampering claims remain contested. What the day establishes for the record is that the physical evidence underpinning several high-profile killings passed through a forensic pipeline the commission now has good reason to scrutinise closely.

Analysis is the archive’s own comment on the public proceedings, grounded in the record above. It weighs the evidence and the process; questions of individual guilt are for the courts and the commission to decide. Everyone named is presumed innocent.